*gwh in Gmc.

[Leo A. Connolly]

petegray wrote:

> There seem to be so many good counter-examples, where *gwh > Germanic w,
> that I confess I am surprised the alternative view exists.  Is there perhaps
> a specific set of circumstances in which it is alleged that *gwh > b, or is
> an idea that can be safely scrap-heaped?

> In addition to snow, warm, and kidney, I find, for example:
> *gwhen  Skt hanmi  Gothic winne (pain) or wunns and the probably related
> English "wound" < *gwhntis

Without a Gothic dictionary at hand, I can only point out that the
parallel forms in other Germanic languages, such as OHG _wunna_, OE
_wyn_ 'joy' and OHG _wini(g)a_ 'Geliebte, wife' belong to PIE *_w(e)n-_,
seen also in Latin _Venus_, the famous goddess.  _Wound_ is plausibly
connected with Gk. _phonos_ 'blood, gore' < *_gwhon-_ but would reflect
the weak grade *_gwhn-to-_ (not **_gwhn-ti-_, else OE **_wynd_), where
syllabic _n_ yields Gmc. /un/.  The words mentioned would suggest that
PIE *gwh eventually yields Gmc. /w/ medially (snow, German Niere) or
before PIE o (Gmc. a) or Gmc. u of any origin, but not before _e_.

> *gwhemer  "swelling"  Latin femur, OE wenn "tumour"

Unlikely -- nn < *mr?

> *gwhoksos  Greek phoksos, OHG wahs "sharp"

The actual OHG form is _hwas(s)_ (OE _hwæs 'sharp'. Goth. _hwassaba_) <
PIE *_kwod-to-_, cf. German _wetzen_, Eng. _whet_ < *_kwod-eje/o-_.

> *gwhesl  "gall"  English weld (conjectured Gothic wizdils, based on late
> Latin uisdil-)

Semantically unacceptable, whatever the origin of _weld_.  And then, why
_gall_ with /g/?

> and some others.

Such as...

> Why is this outcome questioned?

Have to ask Seebold, but one reason is that it explains some Germanic
forms with be- that are otherwise difficult to handle.

Leo